Shows the magnitude of the heating effect on the laminar separation of a water boundary layer and compares its effectiveness in controlling separation to that of surface suction. An analysis is presented of the minimum surface overheat that will delay separation for a prescribed adverse pressure gradient in water. It is found that the effect of heating and viscosity variation is localized to a thin layer near the wall, well within the entire momentum boundary layer. However, although heating is shown to delay separation, the magnitude of its effect is found to be small, particularly if practical values of wall overheat are considered. Results suggest that the effects of surface heating on the delay of separation are probably real and significant, but heating does not have the almost unlimited capacity for maintaining an attached flow that suction appears to possess.